1. Introduction
The sustainability crisis in most Western healthcare systems comprises ever-rising costs (not just in absolute terms, but as a percentage of the gross domestic product) caused by changing demographics and new and often expensive care products in combination with critiques on the quality of care from practitioners and care consumers. In general, healthcare faces a problem of diminishing returns, where the cost increase does not necessarily equate to growth in care quality [
1,
2]. “Sustainability”, a fluid and broad term, then refers to development that “meets the needs of the present generation without compromising the ability of future generations to meet their own needs” [
3] (p.43). The question is whether current healthcare systems will still be able to provide accessible high-quality services for future generations. Sustainability in that sense is a balancing act of different needs and perspectives. Essink [
4] (p.197) frames this as a “dynamic equilibrium in which the system easily rests on its constituent values (e.g., affordability, accessibility, acceptability and quality)”. By following the transition management and system innovation literature [
2,
5], unsustainability can be conceived of as friction between “landscape” developments [
6] and the “regime” [
7]. The dominant institutions and structures have become unequipped to face current societal problems. Diminishing returns and negative side effects are corollaries of system success [
8]. The well-known Multi-Level Perspective (MLP) [
6,
9] can be used to describe how the initial success of a social system drives changes on the landscape level, leading to a misalignment between the successful regime and the needs of the population. It has provided some guidance for prescriptive modeling of sociotechnical transitions [
5]. Although it has been criticized, for instance, for lacking a distinction between “radical” change and regular systemic renewal [
10] and an overemphasis on structures [
11], the model is helpful for eliciting processes of mis- and realignment between the landscape and regime. The model dissects (sociotechnical) systems in three layers: the landscape (societal structures that emerged over multiple generations, like globalization), the regime (structural layer that constitutes the context of common practice with its dominant institutions, physical infrastructure, and culture), and the niche level (protective spaces in which organizers of “initiatives” that deviate from the regime focus on solving problems they recognized on the regime level). Following the MLP, a transition is a regime change that can occur due to sustainability problems related to developments at the landscape level the regime cannot handle, instigated by deviating practices in niches [
5].
Addressing misalignment, and thus systemic unsustainability, through niche innovations is not straightforward. The problems underlying systemic unsustainability are seen as wicked [
12], unstructured [
13], or persistent [
8], meaning the regime actively works against sustainability solutions. To address these problems, support niches, and overcome systemic barriers, the system innovation literature argues for applying reflexivity or, more precisely, a set of methods to enhance reflexivity in (forms of) experimental collaborative practices [
14]. Promoting reflexivity within such initiatives is argued to embed solutions to sustainability problems in existing systemic contexts [
15].
Across health care contexts, the need for reflexivity is increasingly recognized. It has been shown, for instance, that reflexivity at an individual level is an essential part of diagnostic knowing in general practice [
16]. This has led to an increased emphasis on including reflexivity in health profession education (e.g., [
17,
18]). Moreover, the role of team reflexivity in achieving improvements to care provision (e.g., [
19,
20,
21]) and health care innovation (e.g., [
22,
23]) is being explored in an increasing number of settings. Studies that explore the role of reflexivity within transformation of the health care system (i.e., beyond the individual and team level) are, however, sparse.
Furthermore, as this paper will argue, addressing sustainability problems through embedding solutions with the help of reflexive practices cannot equate to “true” sustainability. Even if misalignment is addressed, the implemented, embedded “solutions” will ultimately effectuate new side effects and misalignment. We therefore make a distinction between first-order sustainability, a system in which current sustainability problems are addressed, and second-order sustainability, a system in which the structures are in place to monitor and adapt to misalignment continuously. For such a system, reflexivity itself, as a practice of internal (organizational) accountability [
24], needs to become embedded as a form of structured flexibility. This implies that reflexivity itself needs to become embedded through habitualized ways of thinking and acting [
25] supported by institutional structures [
8].
The aim of this paper is to provide insight into the possibilities of embedding reflexivity into institutions to support second-order sustainability by reporting on two consecutive participatory research programs (6 years of research in total) in different health care contexts where both sought to address unsustainability in terms of misalignment and diminishing returns.
In
Section 2, we explore sustainability issues in maternal care and medicine development in terms of diminishing returns and misalignment, introducing the rationale for interventions based on reflexive methods. Building on that, we argue that reflexivity itself needs to be institutionalized for sustainable systems because of the complexity of sustaining success in addressing sustainability issues. To define our analytical lens, we introduce three main phases of institutionalization: habitualization, objectification, and sedimentation.
In the Results section, we analyze the challenges and opportunities faced in building the “set of methods to enhance reflexive thinking” among the participants in response to sustainability threats. Both programs aimed to facilitate (1) processes of reflection among the project participants on an individual level and (2) the creation of structures to embed these processes sustainably in organizational networks in order to deal with problems of diminishing returns and misalignment. Through this work, we sought to gain insight into the challenges and opportunities of institutionalizing reflexivity.
3. Materials and Methods
We report on the challenges and opportunities faced when we attempted to institutionalize reflexivity in two consecutive participatory research programs, in which the aim was to address unsustainability in terms of misalignment and diminishing returns. In both programs, we collaboratively designed practices with regime actors through reflexive methodologies in order to enhance embedding of both the niche innovation and the reflexive element. The first is the (policy-driven) transformation of perinatal care into a more integrated model with high continuity of care. The second is the European IMI-PARADIGM project aimed at supporting the implementation of meaningful patient engagement in medicine development.
In the two cases, the participatory research approach of Reflexive Monitoring in Action (RMA), as developed by Van Mierlo et al. [
14], was applied both as a research methodology and as a template for institutional structures that support reflexivity. We used this interventionist approach to reflexively monitor the formation of the obstetric partnerships, co-develop a monitoring and evaluation tool within PARADIGM, and develop RMA-based tools to support professionals to work reflexively after the program ends. RMA emphasizes the importance of integrating reflection in the process and promotes reflexive governance, as it encourages actors to scrutinize and reconsider their underlying assumptions, institutional arrangements, and practices in order to steer toward sustainability [
48]. Next to that, RMA aims to stimulate system learning, in which actors learn to (1) recognize the “wickedness” of recurrent problems, (2) acknowledge the systemic barriers but redefine them into opportunities, and (3) design activities that can contribute to systemic change [
49].
Reflexive monitoring follows the circle of observation, analysis, reflection, and adjustment of system innovation initiatives. Research activities undertaken by the monitor include interviews, organizing meetings, and analyzing notes from the meetings and transcripts from the interviews in order to guide the reflection sessions in which action plans are adjusted [
14,
50,
51]. For this paper, we analyzed the transcripts, observation notes, recordings, meeting reports, action agendas, and other materials we gathered in our role as the monitor in order to elicit occurrences of reflexive thinking (or explicit rejection) and the level of habitualization and objectification of this way of working, as described per case below. This analysis thus focused on both the occurrences of reflexive learning on an individual level as well as on the contextual (cultural and institutional) reasons and factors that explain (the lack of) reflexive learning by participants in both projects. We have published elsewhere the results of the programs in terms of integrated care and impact of patient engagement [
26,
31,
52,
53], while in this paper, we focus on the overarching goal of institutionalizing reflexivity.
We used theories on institutionalization and reflexivity to recognize instances of reflexive learning in relation to institutional structures. Themes were drawn around enabling and hampering factors underlying embedding of reflexivity during the two processes in which we employed the principles of RMA, including the relationship of facilitated project activities to wider institutional practices (impact or transformative aspect) and the relations between reflexive practices and institutional resources (i.e., standards).
In the Results section, we present the case analyses separately in order to maintain (1) the logic of how (stages of) case one informed the subsequent stages and case two and (2) to improve the internal validity of the analysis.
3.1. Case 1: Supporting Obstetric Partnerships to Reflect on Perinatal and Client-Centered Care
From 2014 to 2016, we conducted the action research study “North West Netherlands Aligned” (In Dutch: Noordwest Nederland op één lijn), which supported system innovation toward an integrated perinatal and maternal care system by enhancing the collaboration and organizational integration of birth care professionals within obstetric partnerships, with a special focus on client-centered care. The research team, which included two of the authors (T.J. Schuitmaker-Warnaar and J.E.W. Broerse), supported obstetric partnerships for 1 year to build a shared vision on optimal perinatal care and concrete strategies and plans to reach these objectives through attaining short-term goals. The study was embedded within the Maternity Care Network Northwest Netherlands (MCNNN), a regional organization of obstetric partnerships.
The MCNNN is the largest consortium in the Netherlands with around 20% of the national total of births, and at the time, it consisted of 18 active obstetric partnerships, which were all included in this study. Maternity care assistants, primary care and clinical midwives, and obstetricians were always present in meetings and reflection sessions. Some partnerships invited other professionals, like nurses, pediatricians, residents, general practitioners, or youth healthcare professionals. Interviews were held with partnership members, who were selected in consultation with the coordinator of the MCNNN in order to assure that all partnerships and professions would be represented. We held 73 semi-structured interviews and 7 questionnaires among professionals (2 in the whole region in 2015 and 2016 and 5 in specific partnerships as part of the RMA), investigating the desired form of (and barriers and facilitators to) integrated care. Based on the analyses, we organized 18 reflection sessions, with 5 partnerships selected based on geographical spread, implementing the Dynamic Learning Agenda (DLA) [
50,
51,
54]. As for operationalization of the RMA approach [
50], the DLA was used to manage the situated dynamics between different modes of coordination, as it reframes contextual barriers to transformation (toward the intended direction) into learning goals and solutions to overcome them. A DLA explicitly requires reflection from participants as it asks for (1) reframing a problem into a vision on integrated care, (2) systemic barriers for reaching that vision, (3) which other stakeholders can be included, and (4) in what way to work on a solution and the concrete actions to take, culminating in participants drawing up “learning questions”.
Building on the lessons of the program, the DLA approach was developed into a reflexivity supporting toolbox, and we organized four training sessions with four other partnerships who joined voluntarily to explicitly instruct and guide these partnerships to implement the DLA in their own practice, including the reflexive element.
3.2. Case 2: Building a Tool for Structured Reflexivity in Medicine Development
Within the IMI-PARADIGM project, which took place from March 2018 to November 2020, one of the work packages (WP3) worked on developing a tool to monitor and evaluate “the return on engagement” in medicine development. The overall goal of patient engagement throughout the research and development of medicines (PARADIGM,
https://imi-paradigm.eu/ accessed on 18 October 2021) is to develop safer, more effective treatments for patients closely related to their needs and to deliver them faster and more efficiently. In WP3, the Patient Engagement Monitoring and Evaluation Framework (with metrics) was created in order to support organizations in medicine development in their evaluations of the outcomes and the impact of patient engagement in three key phases of medicine development: research priority setting, clinical trial design, and early dialogues with regulatory authorities such as the European Medicines Agency and HTA bodies. The team responsible for the execution of WP3 included three of the authors of this article (T.J. Schuitmaker-Warnaar, C.J. Gunn, and J.E.W Broerse).
The aforementioned RMA approach [
14] was applied to develop and refine an M&E framework. The researchers’ role was to partner with companies and organizations to facilitate discussions and support early attempts to monitor and evaluate their patient engagement initiatives by identifying and selecting relevant metrics for measuring PE impact. Twenty-four case studies of patient engagement initiatives were followed, which elicited in total 47 interviews and 23 reflection sessions with representatives of the pharmaceutical industry, patient organizations, and health governance organizations. The outcomes of the case studies fed into the development of an overall M&E framework among a working group consisting of partners from these stakeholder groups. As well as monthly 1-hour teleconferences (TCs) between the working group across a 2-year period, this process included two 1-day multi-stakeholder workshops (March 2019 and April 2020) which developed consensus on the final overall framework tool. Additionally, two questionnaires were distributed to the working group and their patient engagement networks, with one focusing on their initiative and its (desired) impact and the other about the suitability of the overall framework tool for measuring PE. The framework was presented at three “patient engagement open forum” events, which also provided input from all stakeholder groups.
4. Results
4.1. Institutionalizing Reflexivity in Perinatal Care on the Level of Practitioners
Through the 18 reflection sessions with the participating obstetric partnerships structured by the Dynamic Learning Agenda (DLA), the professionals habitualized reflexivity with the help of the researchers in their role as a reflexive monitor, guiding the DLA process. The partnerships gained insight into the success factors and barriers for integrated care with “client-centricity” and used these insights to challenge institutionalized but undesirable practices and scrutinize “taken for granted” notions by following the four steps of the DLA approach. These reflexive insights were used to create and implement a structured plan to strengthen the care in their region, improve collaboration, and give substance to the client-centered care.
The researchers guided the process of drafting and acting on a DLA by focusing on the contents and quality of care first—something the professionals largely agreed on—laying the groundwork for implementation of the following reflexive question for actors (concerning internal accountability): “Are we (still) doing it right?” In the sessions, professionals collaboratively placed visions on quality at the center of mind maps, operationalized their perspectives on high-quality care in sub-elements and formulated barriers and possibilities for these elements in practice. In other words, they were asked to specify what would the collaboration look like and what would be the possibilities to organize this, given the fact that the mentioned barriers exist. Professionals then formulated reflexive learning questions like "How can good cooperation in the partnership be achieved through shared responsibilities when there are no clear cooperation agreements and protocols on this contribute to the “protocols mountain”?” (VSV 3, Reflection Session 1) and "How can we as healthcare providers take into account the diversity in backgrounds and personal preferences of the client population, given that we want to (and must) deliver a protocoled care process?" (VSV 4, Reflection Session 1). Based on these and other guiding questions, they formulated action plans such as joint intakes of midwives and gynecologists, accompanying each other in the workplace, joint training, drawing up a birth plan with each client (and acting accordingly as care providers), training for the partnerships to generate more from client conversations and structurally embed input from clients in formal meetings (through interviews, surveys, and a client council). In adjusting the action plans based on the results of these actions, the professionals already reflected on their own practice and standardized, embedded, and sedimented modes of working.
Over the course of the project, obstetric partnerships professionalized in terms of the trust between (levels of) professionals, integration of work routines, and overall organization (two questionnaires (one in 2015 and a follow-up in 2016) evaluated the progress in collaboration). They transformed from one large partnership into a smaller board with a mandate and thematic working groups to continuously evaluate care paths and practices (analysis of evaluative interviews in 2015–2016). The reflection sessions showed that the DLA helped to make clear agreements regarding role division, responsibilities, and overall decision making. These agreements and the layered structure of the obstetric partnerships ensured more involvement of members and increased efficiency and effectiveness.
However, the further institutionalization of reflexivity ran into existing structures. In the interviews and reflection sessions, birth care professionals mentioned feeling (unspoken) tension, hierarchy, and mistrust between members, partly arising from the different views between and within different professional groups. Physiological versus pathological views of pregnancy and birth remained a barrier to interprofessional collaboration. Core groups like boards or working group members of obstetric partnerships often managed to overcome this barrier by applying reflexive methods, but in daily practice—regularly and unexpectedly—it became relevant again among the wider group of professionals in the partnerships. In general, the biggest challenge became national developments and news items on those, agitating professionals withing the partnerships and re-introducing distrust. The reflexive process with a focus on quality of care, which all participants could agree on, became disrupted by a policy focus on financial structures, causing care providers to feel that they were forced to think in terms of their own interests instead of the common interest for the pregnant woman.
Furthermore, the process of institutionalizing the reflexive practice of integrating the client perspective did not reach the level of habitualization. Client centeredness was discussed in all reflection sessions and interviews, and its importance was recognized. Partnerships have set up structures for client consultation and participation, mostly through working groups where clients are invited to provide feedback on care paths or protocols. However, structurally shaping client participation remained difficult, in particular due to the lack of knowledge on and experience in client involvement and the fact that caregivers have to do this on top of their already burdensome workloads, as mentioned in sessions and interviews.
Based on these observations, the researchers created a toolbox called “Gezonde geboortezorg met de dynamische leeragenda” (Healthy birth care with the DLA) [
51] that supports professionals working autonomously on applying the DLA within their own obstetric partnership, including the reflexive element. This could potentially increase both the scope and perceived ownership of the reflexive process to explicitly make the improvement of maternity care a project of the care providers themselves. Following the difficulties described above, we assumed "cold" implementation (simply making available and distributing the toolbox) would not yet offer sufficient tools to allow partnerships to work with the toolbox independently. We thus set up an educational program of 1 year together with ZonMW (as part of funding for implementation of successful programs) and the MCNNN, designed to help obstetric partnerships in the Netherlands to work with the toolbox. In this implementation project, we trained representatives to take up the reflexive monitor function in their own partnership to be able to enhance the continuity and quality of perinatal and maternal care with the support of the toolbox and a website (
www.gezondegeboortezorg.nl accessed on 18 October 2021) with a forum for questions and discussion. The four sessions helped to further develop the toolbox based on the experiences from the field in such a way that the representatives of the obstetric partnerships learned to work with the DLA without external support and integrate this with their regular routines, thus habitualizing the reflexive element of the tool.
The toolbox is still (anno 2021) regularly downloaded from the website and other platforms, although the extent to which the reflexive elements are still being applied is unknown. After the different research phases (the reflexive monitoring of the partnerships, the creation of the toolbox, and the training in the use of the toolbox), the following was concluded to be essential for habitualization:
Organization of several consecutive meetings accelerates the process, because results are periodically evaluated and further developed;
External and impartial guidance remains useful;
Participation in physical meetings is stressful but concurrently motivating because of the interaction and tailor-made solutions;
The use of many interactive, playful working methods increases the yield;
Interaction between members of the same partnership greatly helped continuation after the program;
Accreditation (or another tangible incentive from existing institutions) of meetings is an important extra motivation to participate.
In particular, this last lesson implies that broader institutional support remains useful for objectification.
4.2. Institutionalizing Reflexivity in Perinatal Care on the Regional Level
To support institutionalization of the progress of the obstetric partnerships, the steering committee of the regional Maternity Care Network Northwest Netherlands (MCNNN) was also reflexively monitored through recurrent reflection sessions, in which the research team presented the results of the interviews, questionnaires, and progress of the obstetric partnerships. In line with the idea of “institutionalizing reflexivity”, the steering committee was taken as a research object with the aim to create a reflexive knowledge infrastructure capable of providing support to continuous learning within the obstetric partnerships. The steering committee was evaluated by means of (1) observations of meetings, (2) interviews with members, (3) questionnaires, and (4) action-oriented reflection and brainstorming sessions. Visions on perinatal care, tasks, activities, and points for improvement to enhance reflexivity were discussed. The researchers explicitly invited the members to discuss contents of care and not see the committee as a group of representatives from different echelons. The aim was to create a “knowledge network” centered around learning and further development.
The evaluation concluded that the steering committee, as an overarching multidisciplinary body, made an important contribution within the MCNNN. Region-wide research was facilitated, for which individual partnerships did not have the time and financial scopes. Barriers to good care for the entire region were discussed jointly, and collaboration between birth care providers was emphasized and shaped. Various activities and products supported this, including the regional perinatal consultations, the regional protocols, the app, the website, the newsletters, and the studies.
Even though the birth care providers from different echelons within the steering group worked together constructively, friction still arose following national developments, as happened within the obstetric partnerships. The collaboration was further complicated by the rapid changes of the steering committee members and the national pressure on integrated maternity care with associated full personal professional agendas, which precluded the time and energy spent on the steering committee. Important points for attention in the collaboration were increasing the effectiveness of decisions, limiting the individual burden, and offering the space and opportunities for steering committee members to find each other and contribute ideas so that the members remain motivated.
4.3. Difficulties in Reflecting with Professionals
As described, the program itself was rather successful in stimulating reflection on current practices and re-designing new perinatal care practices based on integrated care. However, it remained difficult to institutionalize reflexivity. Even though core groups like the boards of obstetric partnerships adopted the approach, we acknowledge that (1) the process required significant effort to encourage participating professionals to take an actual reflexive stance; (2) after the researchers left, the DLA was usually not maintained; (3) changes in board members further eroded adoption, as they were not trained by the researchers; and (4) boards have tried to organize reflexive sessions with their wider partnerships but were largely unable to convince their colleagues to continue using the reflexive elements of the DLA, reverting to more practical agendas.
Underlying these impediments, we see several systemic features being reproduced, underlining the wickedness of not only first- but also second-order sustainability. First of all, a lot of professionals do not want to “reflect” on their work. When sessions were organized, several participants actively resisted participating in the exercises. A gynecologist, voicing his aversion against “vague and non-scientific” methods, left a reflection session after stating “I’m not going to write on post-its!” (VSV 1, Reflection Session 1). This particular gynecologist repeatedly asked the researchers what the actual “goal” was, a question posed often by others. The act of reflecting was experienced as “vague” as it (by design) expanded the scope of problems before specifying solutions through the creation of action plans. Interestingly, the researchers noticed that the more educated the professionals were, the more they resisted “vague” work that did not seem to have a clear endpoint. On the other hand, some professionals with professional education classified formulating learning questions as part of the DLA as “too academic”.
In the end, however, the organized sessions did add to a more reflexive stance and were perceived as useful by the participants. One participant remarked in a training session on the DLA that “We have been discussing this issue for over 15 years and this is the first time I know what to do next!” (VSV 5, Reflection Session 3). Devising learning questions is a hurdle that takes effort, which needs a well-trained facilitator. After this first step, actions and barriers are discussed, and “reflecting” becomes tangible, making it more interesting for professionals to participate. This, however, is not synonymous with institutionalized reflexivity.
Institutions forced themselves into niche practices regularly. As described above, the continuous focus on financial integration put pressure on the professionals. Mandatory membership of obstetric partnerships was introduced simultaneously with the message that integrated care and bundled payment were the final goal, diverting attempts to co-design new collaborations to complicated discussions on integration. Basic principles for good interdisciplinary teamwork, like leadership, management, vision, and mutual trust, still required development within the partnerships (e.g., [
26]).
Overall, after the various stages aimed at embedding reflexivity, we concluded that better alignment with extant institutional structures might be beneficial for achieving this. More structures need to be in place to allow for different forms of reflexivity. This idea was taken up and incorporated into the proposal for the PARADIGM research project, in which the aim was to draw on the feature of an institutionalized preoccupation with “return on investments” to build structures that presupposed reflexive thinking.
4.4. Institutionalizing Reflexivity through the PARADIGM Framework Exercise
To institutionalize reflexivity, we sought to align with existing structures. The purpose of the PARADIGM project, as introduced in
Section 2.2, was to develop a range of agreed-upon resources and tools for strengthening patient engagement practice, one of which was a framework of evaluation metrics. Through interviews and monthly TCs, the noticeably “louder” industry actors expressed interest in developing a standardized set of metrics that would conclusively “prove” the value of engagement (phrased as the “return on engagement” in popular industry discourse). In the emerging context of patient engagement, where its “embedding” in drug development is of key interest to initiatives and projects like PARADIGM, interests in showcasing and selecting “best practices” undergirded these interests in developing such “rigorous” assessments of the value of patient engagement. Particularly from an industry perspective, impact metrics for patient engagement were already relatively well conceived at the beginning of the project. During the first “Patient Engagement Open Forum” in 2018, participants articulated the importance of various impact indicators related to clinical trial design, including “lower recruitment time” (of participants to a clinical trial), “more diversity (of trial participants) in recruitment”, “retention (of trial participants) rate”, and “fewer trial protocol amendments”.
While these kinds of perspectives on relevant criteria remained relatively unchanged throughout, a “reflexive approach” would need to pay attention to different interests and values in defining impacts for patient engagement, resulting from different actor commitments to medicine development that could not be simply “aligned” in a universal or objective framework of impact metrics. With this in mind, we attempted to facilitate a process that would highlight the relevance of acknowledging and appreciating different stakeholder values and interests in PE in order to generate a final framework that would both align with the voiced interest in standardized metrics and reflect these multiplicities. We attempted to inscribe reflexivity in a seemingly standardized tool.
The approach taken was to construct a framework through a process of negotiating between different perspectives that are invariably present in PE. This process was organized by developing a “general” framework structure that could be adapted to different individual PE initiatives through a tailoring process, following the structure and methodology of RMA.
4.4.1. Designing the Framework Tool
To align the standards with reflexive thinking, we used the tailoring process to explore the multiplicity of perspectives in naming and framing ‘impacts’ of patient engagement. In other words, the selection of meaningful and feasible metrics for a patient engagement initiative invited participants to reflect on the underlying factors or conditions underlying PE activities and their conceived value. These types of reflection were facilitated by the research team during the testing phase in the construction of each case initiative’s “tailored” framework. In developing these tailored frameworks, the organizers of the initiatives were invited to explain under what grounds certain metrics were “relevant” and “feasible” to their practices based on their overall goals or objectives.
Some cases showed the benefits of the tailoring process in its ability to stimulate reflection on how “impact” may be influenced by engagement contexts, including the different stakeholder interests at play. Developing tailored frameworks under the guidance of the reflexive monitors enabled different levels of reflection in different engagement contexts. Many case studies noted the value of the framework approach in relation to the way it stimulated reflection, such as how the exercise was “Thought provoking with good questions for reflection” (Case 21) and “Useful as a prompt for reflection” (Case 5). Especially in settings where multiple stakeholders could be brought together, the processes of reflection on the different frames that influenced the criteria for PE impact were appreciated by the participants (Case 4). Other partners found additional value in the exercise, as it helped them gain “clarity” over the complex processes of patient engagement in their organizations, enabling them to learn about why patient engagement initiatives are being conducted in the first place.
Some partners found it burdensome to work through the long list of reflective questions in developing a tailored evaluation framework. In several cases, the exercise fell flat when facilitators asked how “contextual factors” might influence an engagement initiative. Here especially, more guidance was sought by partners in understanding how these factors played a role in measuring patient engagement.
Furthermore, while the partners recognized the value of defining metrics using such a “co-creative” approach with different stakeholders, many felt restrained by organizational structures (e.g., a lack of time to dedicate to organize complex evaluation and complicated multi-stakeholder exercises being less pressing than more immediate priorities of business). During the testing phase, feedback from one organization was that “the group discussion version would be preferable but impossible to envision due to time pressures" (Case 1). These issues are linked closely to perceptions of the “complexity” of evaluations, to which the concerns of the partners were often related. In the feedback from the testing phase, it was noted that “most common issues were that [the framework] needed simplifying to suit the time pressures of regular organizational work” (Case 14). The interests of several partners in making evaluations “simpler”, aligned with the concerns of “survey fatigue”, related to the burdens of being requested to complete too many evaluations in professional practice (monthly TC number 12). Furthermore, at an open forum (2019) workshop on PE metrics, one participant suggested the use of the net promotor score (NPS) in order to make “straightforward comparisons” in the evaluation of different PE initiatives which, while yielding some insight into how happy participants were with an engagement initiative, would elicit nothing about the reasons why, which is essential information for reflexive processes.
4.4.2. Implementing the Final Framework Tool
A central issue in habitualizing contextual reflection was thus the tangibility of contexts and the practicality of critically reflecting in routine professional settings. We saw that developing tailored frameworks under the guidance of the reflexive monitor enabled different levels of reflection in different engagement contexts. However, the final tool is intended for use by regime actors without facilitation.
The final, “full” metrics framework described a large list of evaluation metrics (n = 87) that asked users to select which metrics were most meaningful to them based on their own engagement contexts, with the predefined “sets of metrics” based on the objectives of engagement, providing a “rough guide” for selecting metrics that align with particular aims or intentions of conducting patient engagement. Like the case studies, the final tool asked users to select metrics relevant to their own practices from the full list, as opposed to defining a conclusive or universal metrics set. The guidance developed therefore encouraged flexibility in the sense that it required some reflection on organizational contexts in order to produce a meaningful set of metrics for those who are using the framework. Crucially, without this type of reflection, the standalone indicators would certainly feel less “rigorous” than the systematic assessments that many in the project showed interest in developing. While some initial users were enthusiastic about the tool, the complex process of adapting the framework to suit local needs was still experienced as difficult to achieve coherently in practice.
The final tool attempted to inscribe different stakeholder perspectives within the predefined sets of metrics. All sets contained different metrics defined as beneficial for all stakeholders. The final tool also encouraged the inclusion of all relevant perspectives during the process of metrics selection. However, several partners in the working group were still interested in developing a more “global” set of “must-have” metrics (workshop observations), which would reduce the need for contextualized selection by users. When reflecting on the framework’s “implementation”, the participants added that the users needed to be more comprehensively guided on how to measure the impact of PE by specifying the most relevant methods for measuring different metrics. Others suggested that the metrics we had developed should form a “soft guide” for organizations to adopt and adapt where necessary, rather than being a strict standard for the field of PE.
4.5. Conceptions of the Role and the Nature of Patient Engagement Evaluation
The development of the framework tool, which encompassed the (facilitation of) various instances of reflection and learning, showed a slow adaptation of the notion of meaningful engagement, meaning that for engagement to yield value, the inclusion of and reflection on multiple perspectives is essential. By the end of the project, this particular stipulation was widely acknowledged.
Furthermore, shifts were seen in the way that the evaluations of patient engagement could be organized to enhance meaningful or valuable engagement practices. In contrast to the “conclusive” impact evaluations envisaged by some, the project leader toward the end of the project began to advocate the value of organizing evaluation as an “internal monitoring” exercise in order for organizations to track their own progress (in achieving valuable or impactful patient engagement) over time (monthly TC number 15).
The research team advocated throughout that local tailoring processes were “better” when performed collectively, including the values of different stakeholders, as this enabled a reflective process more readily. In congruence with previous findings, reflection was found to be more accessible with the support of deliberate (trained) facilitation, such as during the case studies of the testing phase in this project, where researchers could probe the grounds under which the evaluation metrics would be relevant or feasible to different patient engagement practices. At this stage, it remains to be seen how the final framework will be used (i.e., how the envisaged “reflexive approach” to developing metrics for patient engagement will be adopted, adapted, avoided, or resisted by different medicine development actors). Ongoing insights into these dynamics will tell us more about whether and how the strategy of organizing reflexivity proved productive to patient engagement and medicine development.
4.6. Understanding and Supporting Sustained Forms of Reflexivity
Ultimately, developing PARADIGM’s framework for the monitoring and evaluation of patient engagement was an exercise in compromise, with enough coherence for a multi-stakeholder community in medicine development to have their divergent interests in patient engagement (partly) integrated. This required drawing upon features of the system (in this case, a rhetoric of “return on engagement” and a vernacular of impact metrics) to support the institutionalization of more reflexive approaches. The fact that several of the “final” metrics were already well-defined at the start of the project shows that the task was more about striking a balance between different stakeholder interests and values in developing some forms of shared meaning in patient engagement through a process of objectification. This included the development of a collectively defined framework for evaluation between different actor perspectives in patient engagement.
We sought to include meaningful and sustained processes of reflection (the reflexive element) as part of the institutionalization of patient engagement, which came about through the process of tailoring a general (collectively defined) framework into a specific local PE context. We included actors from all levels of the regime, including regulatory authorities such as the European Medicines Agency and HTA bodies, and aligned with their need to stimulate pharmaceutical companies to include more patient-centered outcome measures to be sent for evaluation next to the regular clinical ones. The developed reflexive standards were thus designed to be taken up by—and thereby enforced through—these regulatory agencies. Although we successfully built a tool for standardized reflexivity, it remains to be seen whether these standards will be taken up by the regime.
5. Discussion
The aim of this paper was to provide insight into the possibilities of embedding reflexivity into institutions to support second-order sustainability. We reported on how we applied reflexive methods in two consecutive participatory research programs, in which the aim was to address unsustainability in healthcare in terms of misalignment and diminishing returns. In these analyses, we focused on the challenges and opportunities faced in building the “set of methods to enhance reflexive thinking” among the participants. Both programs aimed to facilitate (1) processes of reflection among the project participants on an individual level and (2) the creation of structures to embed these processes in organizational networks.
In both cases, we were mostly successful in organizing reflexivity within the practices and embedding standardized reflexivity through institutionalization by aligning with regime elements, which was an exercise in compromise, reaching habitualization and some extent of objectification. In perinatal care we, for instance, drew on tangible incentives from existing institutions, and in PARADIGM, we created a framework with standardized metrics to align with industry needs. We supported professionals to enhance internal accountability [
24] and to monitor underlying values. On an individual level, standardized tools with clear steps proved supportive in inviting participants to reflect on their “regular way of doing” because these drew on the systemic element of project-based and solution-oriented professionals in perinatal care by “standardizing” a reflexive process through the Dynamic Learning Agenda and with the M&E framework by enabling interactive learning and priming stakeholders with a broader problem definition. We were able to create “reflexive standards” [
55], implementing the idea that reflexivity can be part of standardization by explicating where reflexivity is situated and who gets to do the reflecting, as well as what is being reflected upon.
In general, the participants in both projects valued the exercises when performed. In perinatal care, for instance, a participant knew what to do next after 15 years of deliberation, and in PARADIGM, the participants valued the “clarity” the framework brought for complex processes of patient engagement. However, the participants also expressed hesitation in engaging with reflexive questions. This aligns with both the increasingly acknowledged value of reflexivity [
16,
19] as well as with the need for more training in different forms of problem solving, like in the case of shared decision making and supporting doctors to consider financial arguments in care decisions [
56,
57], where reflection on the medical practice is a prerequisite. Resistance to reflecting is recognized in the literature, which emphasizes the complexity of the reflexivity concept and the practical complications to the development of reflexive skills [
58].
Reaching habitualization and some extent of objectification, however, did not lead to sedimentation of the new reflexive structures, even though we actively worked on the necessary organizational networks by having relevant regime actors participate in the projects. The willingness of actors to reflect on their daily activities is essential, but the space to maneuver is largely determined by existing institutions. An important systemic feature of the health research system—the results-oriented medical culture—impedes the willingness and ability of participants’ reflexivity, and alignment with contextual factors to enhance the success of system-innovative initiatives is not enough for institutionalization [
38]. In the perinatal care case, the political pressure made professionals less inclined to reflect teamwise. One core assumption of the RMA approach is that focusing on contents and goals helps to overcome strategic conflicts [
14], but this has not been thoroughly examined empirically, and our findings suggest more research is needed on strategies to deal with political conflicts disrupting the reflexive process.
In the PARADIGM case, the participants mentioned, for instance, a lack of time to dedicate to organizing complex evaluation as an example of limited space to maneuver, or that complicated multi-stakeholder exercises are less pressing than the more immediate priorities of businesses. Boström et al. [
59] underscored that reflexive learning and continuous reform is challenging because institutional structures tend to reproduce themselves, unless they are self-confronted by the side effects of their own operations. In PARADIGM, we not only focused on professionals as change agents but explicitly aligned with (the needs of) regime actors, including regulatory authorities, following the lessons from case one and organized a “second-generation initiative”, where the closeness of the regime can function as leverage to scale up niche experiments [
15]. The developed reflexive standards are designed to be taken up by, and thereby enforced through, these regulatory agencies.